Automation in Fund Administration: Why Measurement Has to Come First

A guide for fund accounting, custody, and NAV operations teams

Automation is one of the most widely discussed topics in fund administration operations. It is also one of the most inconsistently applied. Many organisations invest in automation tools and see localized efficiency gains — only to discover that manual effort has been redistributed rather than eliminated, or that the processes targeted for automation were not the ones creating the most operational burden.

The root cause of this pattern is consistent: automation decisions are made based on what is visible or what technology vendors promote, rather than on what operational data reveals about how work actually flows. Without that evidence, even well-resourced automation programmes risk reinforcing existing inefficiencies or creating new ones.

This page examines why measurement has to come before automation in fund administration, where manual effort most commonly concentrates, and how fund accounting and custody teams can use operational data to target automation where it will have the greatest impact — and validate that it is delivering as intended.

Automation initiatives in fund administration tend to follow a recognisable pattern. A technology platform is selected, processes are digitised, and initial results show improvement. Over time, however, the gains prove difficult to sustain. Manual effort reappears in adjacent steps, exception handling becomes more complex, and the operational load on experienced staff does not reduce as expected.

This outcome is not usually the result of poor technology selection. It reflects a more fundamental challenge: the processes selected for automation were not chosen on the basis of where effort actually concentrates, but on where automation was most technically feasible or most visible to management.

Research and industry practice both support this observation. Institutions that approach automation as an operating model discipline — beginning with measurement and evidence — consistently outperform those that treat it as a technology initiative. The difference is not in the tools.

Before any automation decision in fund administration can be made with confidence, three questions need to be answerable with data:

First: where is effort actually concentrating? Not where managers believe it concentrates, and not where system logs indicate activity — but where people are spending time when exceptions, controls, and manual judgement are applied in practice.

Second: which processes are genuinely suitable for automation? A process that is poorly designed or inconsistently executed will not improve when automated. It will fail faster and more expensively. Identifying automation candidates requires visibility into process variability, exception rates, and rework loops — not just task volumes.

Third: is automation reducing effort, or redistributing it? Without post-implementation measurement using the same operational baseline established before automation, organisations cannot distinguish genuine improvement from effort displacement. Automation that appears successful on one metric may be increasing manual workload downstream.

Operational decision intelligence — the continuous measurement of how work flows across people, processes, and systems — provides the foundation for answering all three questions. It creates the evidence base that makes automation decisions defensible and their outcomes verifiable.

Operational measurement across fund administration environments consistently reveals a common distribution of effort. While specific patterns vary by organisation, three categories typically account for the largest share of manual work:

For most fund administration and fund accounting operations, reconciliation represents the clearest case for targeted automation. It involves high-volume, repetitive processes; it produces measurable outputs; and the time consumed by manual reconciliation is typically large enough that even partial automation delivers significant capacity release.

But reconciliation automation is most effective when it is preceded by two steps that are frequently skipped: measurement and centralisation.

These gains were achieved without compromising the accuracy and control standards that fund administration demands — a critical consideration that distinguishes measurement-led automation from technology-led automation.

One of the most overlooked aspects of automation in fund administration is post-implementation measurement. Once automation is in place, organisations frequently move on to the next initiative without systematically confirming whether the intended outcomes have been achieved.

This creates a risk known as automation drift — where automated processes gradually diverge from their intended design, manual workarounds reappear, and the original efficiency gains erode without detection. Because the automation appears to be functioning, the underlying deterioration is not visible until it manifests as a service or capacity problem.

Preventing automation drift requires using the same operational measurement capability applied before automation to assess outcomes after implementation. At CIBC Mellon, post-implementation analysis of task distribution, capacity utilisation, and time allocation confirmed that reconciliation effort had genuinely reduced — and identified where further optimisation was possible.

Automation initiatives in fund administration are more likely to succeed when employees are involved in their design, equipped with relevant skills, and able to see how technology supports rather than threatens their professional development. This is not a peripheral consideration — it directly influences adoption rates, sustainability, and the ability to extract long-term value from automation investment.

At CIBC Mellon, as reconciliation processes were centralised and automated, roles within fund accounting evolved from transaction-oriented execution toward analysis, exception management, and insight generation. Programmes including Skill Sprints and internal hackathons provided structured opportunities for employees to develop digital and analytical capabilities while contributing directly to operational improvement.

The outcome was an organisation where automation was positioned as a tool for empowerment rather than displacement. CIBC Mellon’s most recent employee engagement data reflects this: a 90% employee engagement rating, 7% above the financial services industry norm according to the WTW employee survey.

The results achieved at CIBC Mellon were not limited to reconciliation. Measurement-led automation, combined with the operational visibility that ActiveOps Decision Intelligence provided across the full fund accounting operating model, enabled a series of interconnected improvements:

These results reflect the compounding effect of operational visibility across the full close cycle. When managers can see where workload concentrates, where exceptions are creating rework, and where automation has genuinely reduced rather than displaced effort, they can make better decisions across every aspect of fund administration operations — not just in the processes directly targeted for automation.

ActiveOps provides the operational measurement capability that makes measurement-led automation possible. By capturing effort and workflow data at the point where work is actually performed — across people, processes, and systems — ActiveOps gives fund administration leaders a continuous, granular view of:

• Where effort concentrates across NAV production, reconciliation, exception management, and close-cycle activities.
• How manual work is distributed across teams, funds, and process steps.
• Where automation has genuinely reduced effort and where it has displaced it. How change initiatives are affecting operational capacity and control in real time.

This evidence base supports every stage of the automation lifecycle: identification of candidates, sequencing of interventions, validation of outcomes, and ongoing monitoring to prevent drift.

For technology and transformation leaders in fund administration, it means automation decisions grounded in operational evidence — and confidence that change is delivering what it promised, not just shifting risk to a different part of the system.

See where manual effort concentrates in your fund administration operations before deciding where to automate.

See where manual effort concentrates in your fund administration operations before deciding where to automate.